Handwriting Recognizer

Background:

Through my IBM certificate I had heard of the many modern applications of machine learning and data science and began to explore those applications online. Through exploring these online through articles and blog posts I stumbled upon Handwriting recognition and immediately became curious and attracted to this facet of machine learning.
I wanted to do a project with handwriting recognition and began the project.

Challenges:

There were several challenges I had encountered while working on this recognizer.
The 1st problem I encountered was the that in online tutorials, although they created and trained a model on MNIST data and evaluated the model on a split data set, it was did not accept individual images, so I had to figure out how to have the model accept individual images and attempt to recognize what digit it is. The 2nd problem was that model didn't save itself, so I had to find a way to save the trained model to avoid having to recompile each and every time which took up to 10 minutes. Lastly I had to find a way to create a GUI for writing digits to feed into the model.
Overall these challenges forced me to understand the code and how it's creating a model to recognize code in order to design a recognizer that takes in live inputs.

Future plans:

After completing what I have currently, I hope to be able to learn how to create a recognizer for handwritten alphabets so that it can recognize written letters as well.